liu xueyang 刘雪杨

liu xueyang 刘雪杨

🏳️‍🌈 CS PhD student @ Georgia Tech

Sitemap

A list of all the posts and pages found on the site. For you robots out there is an XML version available for digesting as well.

Pages

Posts

Feeling pain and lost - A new Stoic reflection on the past year and areas to improve

4 minute read

Published:

I was lucky to be at this academic conference (MICRO’55 in Chicago) two weeks ago and the feeling of pain, anxiety and frustration has struck me, seemingly out of nowhere. But I understand every burst of emotion has to come from somewhere, possibly a place I’ve been trying to hide so hard underneath my consciouness, that I thought it’s going ok but deep down it’s not. I haven’t written much about feelings in English, but I’d like to take this chance to reflect on what I did in my Ph.D. first year. I hope this can be a starting point of my self expression. For anyone who happen to be reading this, I hope my personal experience can help you in some way too.

Better machine learning in drones and resource-limited environments

2 minute read

Published:

With machine learning applications flourishing across different platforms, from large-scale data centers to small Internet of Things (IoT) devices, the need for better performance and efficiency keeps growing. In the 2019 system design contest held by Design Automation Conference (DAC), Prof. Cong Hao and her team won the first place with a special software-hardware co-design approach.

portfolio

Spontaneous Social Dancing

Spontaneous Social Dancing is a web-based learning tool designed to help adults feel more at ease initiating casual partner dancing in social or team settings. Through guided movement prompts, etiquette demos, and adaptive support based on user comfort, the platform encourages joyful physical connection without pressure or performance. By blending embodied learning with simple technology, it fosters confidence, consent, and play in shared space. Individual coursework from Education Technology (CS6460).

publications

Mitigating Timing-Based NoC Side-Channel Attacks With LLC Remapping

Published in IEEE Computer Architecture Letters, 2023

Recent CPU microarchitectural attacks utilize contention over the NoC to mount covert and side-channel attacks on multicore CPUs and leak information from victim applications. We propose NoIR, a dynamic LLC slice selection mechanism using slice remapping to obfuscate interconnect contention patterns. NoIR reduces contention variance by 92.18% and mean IPC degradation due to cache invalidation is limited to 7.38% for SPEC CPU 2017 benchmarks for a 1000-access threshold. While previous defenses focused on redesigning the NoC and routing algorithms, we show that a top-down system-level approach can significantly raise the bar for a NoC security vulnerability with minimal modifications to the NoC hardware.

Download here

Accelerator integration in a tile-based SoC: lessons learned with a hardware floating point compression engine

Published in SC-W 23: Proceedings of the SC 23 Workshops of The International Conference on High Performance Computing, Network, Storage, and Analysis, 2023

Heterogeneous Intellectual Property (IP) hardware acceleration engines have emerged as a viable path forward to improving performance in the waning of Moore’s Law and Dennard scaling. In this study, we design, prototype, and evaluate the HPC-specialized ZHW floating point compression accelerator as a resource on a System on Chip (SoC). Our full hardware/software implementation and evaluation reveal inefficiencies at the system level that significantly throttle the potential speedup of the ZHW accelerator. By optimizing data movement between CPU, memory, and accelerator, 6.9X is possible compared to a RISC-V64 core, and 2.9X over a Mac M1 ARM core.

Download here